Test Model Performance Research Articles

Critical considerations of fecal occult blood tests for colorectal cancer

1University of Dundee, Nethergate, Dundee, DD1 4HN, Scotland, UK *Author for correspondence: r.j.c.steele@dundee.ac.uk The most commonly used strategy in colo­ rectal cancer (CRC) screening pro grams involves the use of stool tests to detect occult blood, and guaiac fecal occult blood tests (gFOBTs) are, to date, the only fecal tests shown to reduce CRC mortality in population­based randomized trials [1]. However, gFOBTs also carry disadvantages. A major concern is that interval cancers account for around 50% of cancers detected in gFOBT screened populations [2] and recent data show that gFOBTs are less sensitive in women than in men, and in both rectal and right­sided cancers when compared with left­sided disease [2,3]. In addition, the test is associated with a high false­positive rate with no neoplasia detected in around half of colonoscopies performed following a positive gFOBT [2,3]. This may be, in part, explained by the fact that gFOBTs are not specific for human hemoglobin (Hb) and are subject to possible dietary interference from, for example, red meat and high­peroxidase fruits and vegetables. As a result, it has been common practice to instruct participants to adhere to dietary restrictions ahead of sample collection, although this may act as a barrier to screening and affect participation rates [4]. However, data from a meta­analysis did not support dietary restrictions with gFOBTs, leading to recommendations that restrictions are abandoned to improve adherence rates [5]. Fecal immunochemical tests (FITs) are now available and are increasingly being used in screening programs. Unlike gFOBTs, FITs are specific for the detection of human Hb, eliminating any potential for dietary interference, and are also more specific than gFOBT for lower gastrointestinal bleeding. In addition, modern FITs generally allow a more convenient method of sample collection, and are associated with better participation rates [6]. Another major advantage of FIT is that automated versions not only eliminate interobserver variability, but also the quantitative nature of these tests allow provision of a measured fecal Hb concentration. Screening program “...the adoption of fecal immunochemical tests in colorectal cancer screening programs has potential to address participation, appropriate positivity rates and the problem of false‐negative results.” EDITORIAL

An approach to identify time consistent model parameters: sub-period calibration

Abstract. Conceptual hydrological models rely on calibration for the identification of their parameters. As these models are typically designed to reflect real catchment processes, a key objective of an appropriate calibration strategy is the determination of parameter sets that reflect a "realistic" model behavior. Previous studies have shown that parameter estimates for different calibration periods can be significantly different. This questions model transposability in time, which is one of the key conditions for the set-up of a "realistic" model. This paper presents a new approach that selects parameter sets that provide a consistent model performance in time. The approach consists of testing model performance in different periods, and selecting parameter sets that are as close as possible to the optimum of each individual sub-period. While aiding model calibration, the approach is also useful as a diagnostic tool, illustrating tradeoffs in the identification of time-consistent parameter sets. The approach is applied to a case study in Luxembourg using the HyMod hydrological model as an example.

Constraining ozone‐precursor responsiveness using ambient measurements

This study develops probabilistic estimates of ozone (O3) sensitivities to precursor emissions by incorporating uncertainties in photochemical modeling and evaluating model performance based on ground‐level observations of O3 and oxides of nitrogen (NOx). Uncertainties in model formulations and input parameters are jointly considered to identify factors that strongly influence O3 concentrations and sensitivities in the Dallas‐Fort Worth region in Texas. Weightings based on a Bayesian inference technique and screenings based on model performance and statistical tests of significance are used to generate probabilistic representation of O3 response to emissions and model input parameters. Adjusted (observation‐constrained) results favor simulations using the sixth version of the carbon bond chemical mechanism (CB6) and scaled‐up emissions of NOx, dampening the overall sensitivity of O3 to NOx and increasing the sensitivity of O3 to volatile organic compounds in the study region. This approach of using observations to adjust and constrain model simulations can provide probabilistic representations of pollutant responsiveness to emission controls that complement the results obtained from deterministic air‐quality modeling.

EVALUATING THE PREDICTIVENESS AND PROFITABILITY OF FOREIGN EXCHANGE RATE FORECASTING MODELS

This paper evaluates the performance of two competing currency models as a forecasting and trading tool in fund management. A dynamic vector error correction model is utilized to construct a currency forecasting and fair value forecasting model for the Euro-Dollar exchange rate. Emphasis is placed on robustness testing model performance by changing its specification and how both models perform across different time periods. Based on the accuracy of the forecasts the fair value model outperforms the currency forecasting model; a finding that is not supported using directional forecasts. This is robust to changes in model specification and across different time spans that cover pre-and current financial crisis periods. It is also discovered that the evaluation criteria used and prevailing market conditions determines whether model performance translates into value added in a currency fund.

Testing the predictive performance of distribution models

Distribution models are used to predict the likelihood of occurrence or abundance of a species at locations where census data are not available. An integral part of modelling is the testing of model performance. We compared different schemes and measures for testing model performance using 79 species from the North American Breeding Bird Survey. The four testing schemes we compared featured increasing independence between test and training data: resubstitution, random data hold‐out and two spatially segregated data hold‐out designs. The different testing measures also addressed different levels of information content in the dependent variable: regression R2 for absolute abundance, squared correlation coefficient r2 for relative abundance and AUC/Somer’s D for presence/absence. We found that higher levels of independence between test and training data lead to lower assessments of prediction accuracy. Even for data collected independently, spatial autocorrelation leads to dependence between random hold‐out test data and training data, and thus to inflated measures of model performance. While there is a general awareness of the importance of autocorrelation to model building and hypothesis testing, its consequences via violation of independence between training and testing data have not been addressed systematically and comprehensively before. Furthermore, increasing information content (from correctly classifying presence/absence, to predicting relative abundance, to predicting absolute abundance) leads to decreasing predictive performance. The current tests for presence/absence distribution models are typically overly optimistic because a) the test and training data are not independent and b) the correct classification of presence/absence has a relatively low information content and thus capability to address ecological and conservation questions compared to a prediction of abundance. Meaningful evaluation of model performance requires testing on spatially independent data, if the intended application of the model is to predict into new geographic or climatic space, which arguably is the case for most applications of distribution models.

Design optimization of a high specific speed Francis turbine runner

Francis turbine is used in many hydroelectric power stations. This paper presents the development of hydraulic performance in a high specific speed Francis turbine runner. In order to achieve the improvements of turbine efficiency throughout a wide operating range, a new runner design method which combines the latest Computational Fluid Dynamics (CFD) and a multi objective optimization method with an existing design system was applied in this study. The validity of the new design system was evaluated by model performance tests. As the results, it was confirmed that the optimized runner presented higher efficiency compared with an originally designed runner. Besides optimization of runner, instability vibration which occurred at high part load operating condition was investigated by model test and gas-liquid two-phase flow analysis. As the results, it was confirmed that the instability vibration was caused by oval cross section whirl which was caused by recirculation flow near runner cone wall.

A framework for benchmarking land models

Abstract. Land models, which have been developed by the modeling community in the past few decades to predict future states of ecosystems and climate, have to be critically evaluated for their performance skills of simulating ecosystem responses and feedback to climate change. Benchmarking is an emerging procedure to measure performance of models against a set of defined standards. This paper proposes a benchmarking framework for evaluation of land model performances and, meanwhile, highlights major challenges at this infant stage of benchmark analysis. The framework includes (1) targeted aspects of model performance to be evaluated, (2) a set of benchmarks as defined references to test model performance, (3) metrics to measure and compare performance skills among models so as to identify model strengths and deficiencies, and (4) model improvement. Land models are required to simulate exchange of water, energy, carbon and sometimes other trace gases between the atmosphere and land surface, and should be evaluated for their simulations of biophysical processes, biogeochemical cycles, and vegetation dynamics in response to climate change across broad temporal and spatial scales. Thus, one major challenge is to select and define a limited number of benchmarks to effectively evaluate land model performance. The second challenge is to develop metrics of measuring mismatches between models and benchmarks. The metrics may include (1) a priori thresholds of acceptable model performance and (2) a scoring system to combine data–model mismatches for various processes at different temporal and spatial scales. The benchmark analyses should identify clues of weak model performance to guide future development, thus enabling improved predictions of future states of ecosystems and climate. The near-future research effort should be on development of a set of widely acceptable benchmarks that can be used to objectively, effectively, and reliably evaluate fundamental properties of land models to improve their prediction performance skills.

Evaluation of climate models using palaeoclimatic data

There is large uncertainty about the magnitude of warming and how rainfall patterns will change in response to any given scenario of future changes in atmospheric composition and land use. The models used for future climate projections were developed and calibrated using climate observations from the past 40 years. The geologic record of environmental responses to climate changes provides a unique opportunity to test model performance outside this limited climate range. Evaluation of model simulations against palaeodata shows that models reproduce the direction and large-scale patterns of past changes in climate, but tend to underestimate the magnitude of regional changes. As part of the effort to reduce model-related uncertainty and produce more reliable estimates of twenty-first century climate, the Palaeoclimate Modelling Intercomparison Project is systematically applying palaeoevaluation techniques to simulations of the past run with the models used to make future projections. This evaluation will provide assessments of model performance, including whether a model is sufficiently sensitive to changes in atmospheric composition, as well as providing estimates of the strength of biosphere and other feedbacks that could amplify the model response to these changes and modify the characteristics of climate variability.

Non-singular slip (NSS) method for longitudinal tire force calculations in a sudden braking simulation

In a dynamic vehicle simulation, longitudinal tire force is primarily based on the longitudinal slip (ratio). In the longitudinal slip formula, state variables are used in the denominator. This causes a divergence problem for numerical simulations of vehicle dynamics. To avoid this numerical singularity, a differential slip calculation method was developed for use in dynamic vehicle simulations. However, this method also causes a singularity when the wheel velocity approaches zero in a pure slip state, such as during sudden braking. In this paper, a new longitudinal slip calculation method, which can overcome singularities in all velocity conditions, is proposed. For this purpose, the Taylor series is adapted to the slip formula and the idea of virtual wheel rotation stiffness is introduced for the development of the slip equation. The physical phenomenon at the zero slip state is analyzed. Finally, the proposed slip formula is used to solve the numerical singularity problem, and the non-singular slip (NSS) calculation method is proposed. The proposed NSS method is applied to tire model performance test (TMPT) simulations to validate its performance.

Insights into North Pacific right whale Eubalaena japonica habitat from historic whaling records

ESR Endangered Species Research Contact the journal Facebook Twitter RSS Mailing List Subscribe to our mailing list via Mailchimp HomeLatest VolumeAbout the JournalEditorsSpecials ESR 15:223-239 (2011) - DOI: https://doi.org/10.3354/esr00381 Theme Section: Beyond marine mammal habitat modeling: applications for ecology and conservation Insights into North Pacific right whale Eubalaena japonica habitat from historic whaling records Edward James Gregr* SciTech Environmental Consulting, 2136 Napier Street, Vancouver, British Columbia V5L 2N9, Canada *Email: ed@scitechconsulting.com ABSTRACT: Whaling records from the mid-1800s provide the largest set of observations with which to conduct a basin-scale analysis of potential North Pacific right whale Eubalaena japonica habitat. Since these data lack the concurrent oceanographic data necessary to investigate the species’ habitat characteristics I used ocean climate from a 20th century circulation model to create a suitable set of habitat predictors. My goals were to (1) identify regions of suitable habitat and (2) investigate the processes underlying the species–habitat relationship by (3) examining model performance at different spatial and temporal scales. The results show 2 non-overlapping habitat regions in the subarctic North Pacific, supporting the notion of 2 distinct subpopulations. The analysis also implicates surface temperature and temperature variability as strong indicators of potential right whale habitat. Tests of model performance at different scales strongly suggest that at the basin-scale, right whales use regions of cold water with low inter-annual variability and high within-season variability (i.e. areas where high frontal activity occurs predictably from year to year). The significance of these indicators decreased at the regional scale emphasising the coupling of scale and process, and thus the need for different predictors at different scales. Comparisons of models built using different subsets of the dependent data showed how hypotheses can be tested and potential biases in observational data can be explored. Analyses of rare species’ habitat such as this can provide guidance for more directed survey efforts and help identify areas and processes of potential biological importance. KEY WORDS: Habitat suitability model · Pelagic habitat · Concentration features · Critical habitat · Habitat processes · Ecological niche · Presence-only · Maximum entropy Full text in pdf format PreviousNextCite this article as: Gregr EJ (2011) Insights into North Pacific right whale Eubalaena japonica habitat from historic whaling records. Endang Species Res 15:223-239. https://doi.org/10.3354/esr00381 Export citation RSS - Facebook - Tweet - linkedIn Cited by Published in ESR Vol. 15, No. 3. Online publication date: December 12, 2011 Print ISSN: 1863-5407; Online ISSN: 1613-4796 Copyright © 2011 Inter-Research.

A parsimonious geomorphological unit hydrograph for rainfall–runoff modelling in small ungauged basins

In this study, a parsimonious hydrological modelling algorithm is proposed based on the automated DEM-based geomorphic characterization of runoff dynamics in scarcely monitored river basins. The proposed approach implements the instantaneous unit hydrograph (IUH) concept, estimated using the width function (WF), for characterizing the travel time distribution using just one parameter, the river network flow velocity. Hillslope velocities are defined using spatially-distributed empirical formulas based on slope and soil-use information extrapolated from digital topographic data. Case studies are presented for testing model performance and comparing simulated and observed hydrographs of 25 selected flood events, as well as investigating the differences with the geomorphological instantaneous unit hydrograph (GIUH) model results. The calibration of the WFIUH channel flow velocity parameter using the concentration time is investigated providing interesting insights for the use of such a method for hydrological prediction in ungauged basins. Editor D. Koutsoyiannis Citation Grimaldi, S., Petroselli, A. and Nardi, F., 2012. A parsimonious geomorphological unit hydrograph for rainfall–runoff modelling in small ungauged basins. Hydrological Sciences Journal, 57 (1), 73–83.

Effects of dynamically variable saturation and matrix‐conduit coupling of flow in karst aquifers

Well‐developed karst aquifers consist of highly conductive conduits and a relatively low permeability fractured and/or porous rock matrix and therefore behave as a dual‐hydraulic system. Groundwater flow within highly permeable strata is rapid and transient and depends on local flow conditions, i.e., pressurized or nonpressurized flow. The characterization of karst aquifers is a necessary and challenging task because information about hydraulic and spatial conduit properties is poorly defined or unknown. To investigate karst aquifers, hydraulic stresses such as large recharge events can be simulated with hybrid (coupled discrete continuum) models. Since existing hybrid models are simplifications of the system dynamics, a new karst model (ModBraC) is presented that accounts for unsteady and nonuniform discrete flow in variably saturated conduits employing the Saint‐Venant equations. Model performance tests indicate that ModBraC is able to simulate (1) unsteady and nonuniform flow in variably filled conduits, (2) draining and refilling of conduits with stable transition between free‐surface and pressurized flow and correct storage representation, (3) water exchange between matrix and variably filled conduits, and (4) discharge routing through branched and intermeshed conduit networks. Subsequently, ModBraC is applied to an idealized catchment to investigate the significance of free‐surface flow representation. A parameter study is conducted with two different initial conditions: (1) pressurized flow and (2) free‐surface flow. If free‐surface flow prevails, the systems is characterized by (1) a time lag for signal transmission, (2) a typical spring discharge pattern representing the transition from pressurized to free‐surface flow, and (3) a reduced conduit‐matrix interaction during free‐surface flow.

Diagnostic performance of quantitative fecal immunochemical test and multivariate prediction model for colorectal neoplasms in asymptomatic individuals

Quantitative fecal immunochemical test (QTFIT) has the advantage of being able to describe test characteristics on a scaled rather than binary system. The aims of this study were to decide the optimal cut-off points of QTFIT and to make a multivariate prediction model for colorectal neoplasms in asymptomatic adults. We retrospectively analyzed 1085 consecutive asymptomatic individuals who completed both full colonoscopy and QTFIT at a general health checkup clinic. Advanced adenomatous polyps (AP) were defined as APs of at least 1 cm in diameter; adenomas with villous component or high-grade dysplasia; and significant neoplasia (SN) including both advanced AP and colorectal cancer. The ideal cut-off value of QTFIT was chosen based on a value that maximized the sum of both sensitivity and specificity, and clinical utility. For AP, 25 ng/ml was chosen as the optimal cut-off value and provided a sensitivity of 31% [95% confidence interval (CI): 27-36] and specificity of 79% (95% CI: 76-82). For SN, the ideal QTFIT cut-off value was 25 ng/ml, providing a sensitivity of 51% (95% CI: 39-62) and specificity of 77% (95% CI: 74-80). For colorectal cancer, the optimal cut-off point was 50 ng/ml, offering a sensitivity of 75% (95% CI: 41-93) and specificity of 86% (95% CI: 85-86). The multivariate prediction model was represented by nomogram and was validated by bootstrap method. The diagnostic performance of QTFIT for CRC is promising, although its sensitivity for AP and SN is unsatisfactory. BMI, in addition to age and sex improves the accuracy of SN screening by QTFIT.

A provider-based water planning and management model – WaterSim 4.0 – For the Phoenix Metropolitan Area

Uncertainty in future water supplies for the Phoenix Metropolitan Area (Phoenix) are exacerbated by the near certainty of increased, future water demands; water demand may increase eightfold or more by 2030 for some communities. We developed a provider-based water management and planning model for Phoenix termed WaterSim 4.0. The model combines a FORTRAN library with Microsoft C# to simulate the spatial and temporal dynamics of current and projected future water supply and demand as influenced by population demographics, climatic uncertainty, and groundwater availability. This paper describes model development and rationale. Water providers receive surface water, groundwater, or both depending on their portfolio. Runoff from two riverine systems supplies surface water to Phoenix while three alluvial layers that underlie the area provide groundwater. Water demand was estimated using two approaches. One approach used residential density, population projections, water duties, and acreage. A second approach used per capita water consumption and separate population growth estimates. Simulated estimates of initial groundwater for each provider were obtained as outputs from the Arizona Department of Water Resources (ADWR) Salt River Valley groundwater flow model (GFM). We compared simulated estimates of water storage with empirical estimates for modeled reservoirs as a test of model performance. In simulations we modified runoff by 80%–110% of the historical estimates, in 5% intervals, to examine provider-specific responses to altered surface water availability for 33 large water providers over a 25-year period (2010–2035). Two metrics were used to differentiate their response: (1) we examined groundwater reliance (GWR; that proportion of a providers’ portfolio dependent upon groundwater) from the runoff sensitivity analysis, and (2) we used 100% of the historical runoff simulations to examine the cumulative groundwater withdrawals for each provider. Four groups of water providers were identified, and discussed. Water portfolios most reliant on Colorado River water may be most sensitive to potential reductions in surface water supplies. Groundwater depletions were greatest for communities who were either 100% dependent upon groundwater (urban periphery), or nearly so, coupled with high water demand projections. On-going model development includes linking WaterSim 4.0 to the GFM in order to more precisely model provider-specific estimates of groundwater, and provider-based policy options that will enable “what-if” scenarios to examine policy trade-offs and long-term sustainability of water portfolios.

Failure by fatigue in the field: a model of fatigue breakage for the macroalga Mazzaella, with validation

Seaweeds inhabiting the extreme hydrodynamic environment of wave-swept shores break frequently. However, traditional biomechanical analyses, evaluating breakage due to the largest individual waves, have perennially underestimated rates of macroalgal breakage. Recent laboratory testing has established that some seaweeds fail by fatigue, accumulating damage over a series of force impositions. Failure by fatigue may thus account, in part, for the discrepancy between prior breakage predictions, based on individual not repeated wave forces, and reality. Nonetheless, the degree to which fatigue breaks seaweeds on wave-swept shores remains unknown. Here, we developed a model of fatigue breakage due to wave-induced forces for the macroalga Mazzaella flaccida. To test model performance, we made extensive measurements of M. flaccida breakage and of wave-induced velocities experienced by the macroalga. The fatigue-breakage model accounted for significantly more breakage than traditional prediction methods. For life history phases modeled most accurately, 105% (for female gametophytes) and 79% (for tetrasporophytes) of field-observed breakage was predicted, on average. When M. flaccida fronds displayed attributes such as temperature stress and substantial tattering, the fatigue-breakage model underestimated breakage, suggesting that these attributes weaken fronds and lead to more rapid breakage. Exposure to waves had the greatest influence on model performance. At the most wave-protected sites, the model underpredicted breakage, and at the most wave-exposed sites, it overpredicted breakage. Overall, our fatigue-breakage model strongly suggests that, in addition to occurring predictably in the laboratory, fatigue-induced breakage of M. flaccida occurs on wave-swept shores.

Modeling Phosphorus Transport in an Agricultural Watershed Using the WEPP Model

The Water Erosion Prediction Project (WEPP) model has been tested for its ability to predict soil erosion, runoff, and sediment delivery over a wide range of conditions and scales for both hillslopes and watersheds. Since its release in 1995, there has been considerable interest in adding a chemical transport element to it. Total phosphorus (TP) loss at the watershed outlet was simulated as the product of TP in the soil, amount of sediment at the watershed outlet, and an enrichment ratio (ER) factor. WEPP can be coupled with a simple algorithm to simulate phosphorus transport bound to sediment at the watershed outlet. The objective of this work was to incorporate and test the ability of WEPP in estimatingTP loss with sediment at the small watershed scale. Two approaches were examined. One approach (P-EER) estimated ER according to an empirical relationship; the other approach used the ER calculated by WEPP (P-WER).The data used for model performance test were obtained from two side-by-side watersheds monitored between 1976 and 1980. The watershed sizes were 5.05 and 6.37 ha, and each was in a corn (Zea mays L.)-soybean [Glycine max (L.) Merr.] rotation. Measured and simulated results were compared for the period April to October in each year. There was no statistical difference between the mean measured and simulated TP loss. The Nash-Sutcliffe coefficient was 0.80 and 0.78 for the P-EER and P-WER methods, respectively. It was critical for both methods that WEPP adequately represent the biggest sediment yield events because sediment is the main driver for TP loss so that the model can adequately simulate TP losses bound to sediment. The P-WER method is recommended because it does not require use of empirical parameters to estimate TP loss at the watershed outlet.

Nested circulation modelling of inter-tidal zones: details of a nesting approach incorporating moving boundaries

Nested circulation models developed to date either exclude the flooding and drying process or prohibit flooding and drying of nested boundaries; they are therefore ill-suited to the accurate modelling of inter-tidal areas. The authors have developed a nested model with moving boundaries which permits flooding and drying of both the interior domain and the nested boundaries. The model uses a novel approach to boundary formulation; ghost cells are incorporated adjacent to the nested boundary cells so that the nested boundaries operate as internal boundaries. When combined with a tailored adaptive interpolation technique, the approach facilitates a dynamic internal boundary. Details of model development are presented with particular emphasis on the treatment of the nested boundary. Results are presented for Cork Harbour, a natural coastal system with an extensive inter-tidal zone and a complex flow regime which provided a rigorous test of model performance. The nested model was found to achieve the accuracy of a high resolution single grid model for a much lower computational cost.

Use of simulated data from a process‐based habitat model to evaluate methods for predicting species occurrence

Much research has centered on determining which habitat model best predicts species occurrence. However, previous work typically used data sets that are inherently biased for evaluation. The use of simulated data provides a way of testing model performance using un‐biased data where the true relationships between species occurrence and population processes are predefined using sound ecological theory. We used a process‐based habitat model to generate simulated occurrence data to evaluate presence–absence and presence–only methods: generalized linear and generalized additive models (GLM, GAM), maximum entropy model (Maxent), and discrete choice models (DCM). This is the first study to use a DCM for predicting species distributions. We varied the effect that habitat quality had on fecundity and reported the model responses to these changes. When the effect of habitat quality on fecundity was weak, model performance was no better than random for all methods, however, performance increased as the habitat/fecundity relationship became stronger. For each level of habitat quality effect, there was little variation in performance between the presence–absence and presence–only methods. The use of a process‐based habitat model to generate occurrence data for evaluating model performance has a distinct advantage over other testing methods, because no errors are made during sampling and the true ecological relationships between population process and species occurrence are known. This leads to un‐biased results and increased confidence in assessing model performance and making management recommendations.

Ecosystem models clarify the trophic role of whales off Northwest Africa

There is global concern about the interaction between whales and fisheries, and in some countries, great whales are viewed as a threat to fisheries by potentially eating fish species that could be exploited for human consumption. We developed an ecosystem model to explore the trophic inter- actions between cetaceans and fisheries off Northwest Africa and to examine the potential impact of a reduction in the abundance of baleen whales on fishery yields. This allowed us to characterize the structure and function of the ecosystem in terms of biomass, mortalities, consumption rates, food habits, and fisheries. Faced with sparse data for our study area, we explicitly accounted for uncer- tainty in ecosystem structure, model accuracy, and input data and conducted an extensive sensitivity analysis. We tested model performance with time series of biomass and catches for important species of the system. Our results indicate that the overlap between prey species consumed by cetaceans and species targeted in fisheries is low. Furthermore, for a wide range of assumptions about whale abun- dances, diet composition, and food consumption in breeding areas, we found that whale consumption is several orders of magnitude lower than total fishery catches and 2 orders of magnitude lower than the amounts taken by other trophic groups. Finally, simulations of substantial reductions of whale populations did not influence the biomass of commercially important fish, nor any other species of the foodweb. These results suggest that fisheries yields would not benefit from the removal of whales in this area.

A parametric daily precipitation model application in Botswana

A parametric precipitation model is developed for generation of daily rainfall time series based on historic data. The precipitation model is a composite model of Markov-chain (MC) and probability distribution (PD). Thirty nine rain gauge stations in Botswana that have daily rainfall record length in the range of 11 to 83 years have been considered to test the applicability of the model. Results show that Markov-chain (MC) model can be used to model the persistence behaviour of the transition probability matrix (TPM) of dry and wet day rainfall sequences. With the MC model, the two-parameter gamma distribution is found to be most robust and suitable model to describe the magnitude of rainfall depths in wet days. The performance of the gamma distribution was evaluated among other three candidate distributions we considered, namely the lognormal, the exponential and weibull distributions. We have conducted model performance tests of the proposed precipitation model and its MC and PD components with the historical records. Both the sequence of wet-dry day sequences and rain day rainfall magnitudes were simulated with the seasonally varying parameter sets we have developed. Simulated rainfall, both its dry-wet day sequence and rainy day rainfall depths were closely comparable with the historical records with reasonably success rates. Regression coefficient between the simulated and observed annual rainfall at the 39 stations is 66%.Key Words: Markov Chain, Transition Probability Matrix, Gamma distribution, Parameter, Descriptive ability test